Annals of Botany 93: 115, 2004
© 2004 Annals of Botany Company
Skelton P, Smith A. 2002. Accompanying booklet written by Monks N. Cladistics: a practical primer on CD-ROM
Cladistics: a practical primer on CD-ROM.
Skelton P, Smith A. 2002. Accompanying booklet written by Monks N
Cambridge University Press. Co-produced by the Natural History Museum, the Open University and the BBC.
£29·95. pp.
As the authors state in their preface, the ‘last few decades have seen a revolution in the reconstruction of evolutionary relationships’, fuelled, in particular, by the use of powerful computers and the relatively easy availability of DNA sequence data. For beginners and scientists in other fields of biology, this interactive CD-ROM provides an introduction to the basic methodology of cladistics, using examples for both morphology and DNA sequences. The CD-ROM is accompanied by a booklet that provides the text and examples, as well as additional notes and comments. The user does not need to read the booklet, but it is useful to have it to refer to; it helps to find the parts of the computerized presentation to which the user would like to return. Overall, I found this to be an excellent production. I have a minor quibble with the fact that the booklet says that there is a start icon that must be clicked, but in fact there is no start icon at all – however, I soon figured out that I had to click on the program (a standard way to start programs on a Macintosh; the CD operates on both PCs and Macs). There are introductory comments in each of the five sections, and these provide definitions of terminology, which are then followed by examples and problems that the user is given the opportunity to solve interactively. There are parts with just written text and others in which there is a narrator. The first section, First Principles, explains the general idea of cladistics and what it is designed to do. It also covers the principle of parsimony and carefully and accurately explains how it is applied in an evolutionary context. This was a well-organized introduction to the topic, but it would have been nice if the authors had clearly distinguished between ‘phylogeny’ (the true history, in which ancestor/descendent relationships took place and which is ultimately unknowable) and the product of cladistic analysis, which is just an estimate of the phylogeny. I also have to say that I found the pronunciation of ‘parsimony’ and ‘parsimonious’ throughout the spoken sections a little strange; it was as if the words were spelled ‘parzimony’ and ‘parzimonious’.
The second section covers characters and the concept of homology, which is followed by a definition of character types and how one could go about establishing weights to be incorporated into the analysis. There are good examples from morphology as well as gene sequences. In one of the problems to be solved after having had the difference between a ‘variable site’ and an ‘informative site’ explained, the user is given a matrix with four DNA sequences for the ingroup as well those for one outgroup (which of course is established based on other information before analysis commences). The user is told that characters unique to the outgroup are informative, but of course this is not true; unique characters are always uninformative. Although it is true that these autapomorphies distinguish the ingroup from the outgroup, that distinction was already made independently of the analysis itself. An autapomorphy (‘phylogenetically neutral sites’ as termed here) can never provide any structure to a cladogram, and being in the outgroup does not change this. Such characters provide no structure to the tree, and it is as confusing for novices to include them in the same category as those variable in two or more taxa.
The third section describes cladograms and the various types of trees. Throughout the text (written and spoken), there is some confusion over what a node is. There is an important difference between a cladogram, which shows only a hypothetical branching pattern leading to the terminal taxa included in the analysis, and a phylogeny, which includes the branching pattern plus ancestor/descendent relationships. The latter can be the subject of speculation, but it can never been known with certainty. One of the most important distinctions between the two is the concept of node; in the former, a node, as the authors point out, ‘corresponds to nested sets of taxa that are defined by their shared character states’ and ‘strictly speaking they do not represent the latest common ancestral taxa. . .’. In a phylogeny (the real course of events), nodes represent common ancestors. The authors therefore do know the difference, but they then resort to describing a monophyletic group as the latest common ancestor and all of its descendants. If the nodes of cladograms do not represent taxa (last common ancestor), then you cannot classify nodes as part of a clade, which can in reality include only the terminal taxa in the matrix. This distinction may appear trivial, but it is not because it is the basis of the arguments for those who attack cladistic classification as untenable: if nodes are ancestors to be included in named taxa (let us say taxon A), then the next node deeper cannot be included in a different taxon (taxon B, the sister of A) because then some of its descendants (taxon A) are not part of taxon B. If you use a node-based definition of clades, then there is only one clade, all life on the planet. Therefore, I am highly critical of the lack of clarity in this product about the definitions of nodes and what they represent, the definitions of monophyly and paraphyly and the distinction between a cladogram and phylogeny.
The authors also distinguish between a cladogram and a phylogram, the latter of which Dave Swofford used in the PAUP program (phylogenetic analysis using parsimony, one of the most commonly used tree-building softwares) for cladograms that include amount of change as length on the horizontal branches. This is an old term originally used for the output of phenetic analyses, which included all information in the analysis regardless of whether the characters were primitive or derived and which always included branch lengths. A phylogram produced by a phenetic analysis and a tree with branch lengths produced by a parsimony analysis are not the same thing, so it was unwise to use the term phylogram, but at this point its use is so common that making a distinction is probably a lost cause. Similarly, the authors do not clearly distinguish between ‘phylogeny’ and ‘phylogenetic trees’, the former being the real thing and the latter only an estimate of the former. I would argue that, in their example of the rudists, they have illustrated a stratigraphic hypothesis of the phylogeny of this group, which is not a phylogenetic tree because the former includes ancestor/descendent relationships and the latter would not. A cladogram and a phylogenetic tree (or tree for short) are both hypotheses in which nodes represent nested sets of taxa.
The fourth section explains how the clarity or robustness of phylogenetic trees is established. As above, I have a problem with the authors’ description of what is taking place: ‘Finding the most parsimonious cladogram is by no means the end of search for a phylogeny, as questions about its reliability remain’. What they clearly mean here by ‘phylogeny’ is an estimate of the phylogeny of a group because there is no point questioning the ‘reliability’ of what really happened.
The fifth and last chapter is a practical in which the reader is supposed to construct a morphological matrix (which could take more than 4 h) as well as aligning a set of DNA sequences by inserting gaps into the matrix. Bootstrap percentages are provided for each of the two separate analyses, as well as for a direct combination of the two. Apparent points of incongruence are present in the separate analyses, and a brief discussion of how bootstrap percentages change in the combined analysis is an indication of general congruence. The authors do not go into the debates of how best to combine independent data sets and provide only a ‘total evidence’ approach. All of this is excellent practice and worthwhile.
Although I have been highly critical over what I consider to be major flaws of this production, I nonetheless think that it provides the novice with an easily understood introduction to the general topics of how morphological and DNA sequence data are used in parsimony analyses. I give it my somewhat less than whole-hearted approval and hope that these flaws are corrected in the next edition. In spite of my misgivings about how some topics are handled, there is still a great deal of utility in this approach to teaching cladistic analyses, and I would nevertheless give it to my students to help them become familiarized with the topic. There is no other product available to fill this niche. It is particularly useful because of the combined approach of morphological and molecular data.
LITERATURE CITED
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||